This form provides a restricted interface to the luminosity calculator iLumiCalc.exe. A Good Run List appropriate for your processed data sample is required as an input.
Useful Links
- Luminosity Recommendations for Physics
- Good Runs List TWiki
- LumiCalc Tutorial
- Offline Luminosity Tags
- Luminosity Task Force Report
- LAr Event Veto instructions
LumiCalc News
If your request seems to hang and you eventually see a Proxy Error after clicking Calculate Luminosity, this is annoying but not fatal. Just reload the page and your job should be running fine.
- 30 Jul, 2020 - Migration to centos7 server complete
- 22 Jun, 2020 - Fix bug in handling SFO misconfiguration in run 281385
- 16 Apr, 2020 - Preliminary 13 TeV 2017/18 low-mu lumi available in tag OflLumi-13TeV-010-lowmu
- 6 Feb, 2020 - Final 2015 PbPb lumi available in tag OflLumi-HI-009
- 17 Nov, 2019 - Updated 2015 PbPb lumi available in tag OflLumi-HI-008
- 19 Oct, 2019 - Tag OflLumi-5TeV-003 updated with prelim. 2017 5 TeV pp
- 6 Mar, 2019 - Preliminary 2018 PbPb luminosity available in tag OflLumi-HI-007
- 18 Feb, 2019 - Preliminary 2018 pp luminosity available in tag OflLumi-13TeV-010
- 18 Dec, 2018 - Tag OflLumi-HI-006 available
- 12 Nov, 2018 - Tag OflLumi-900GeV-002 available, prelim. 2018 900 MeV pp
- 17 Sep, 2018 - Tag OflLumi-5TeV-003 available, final 2015 5 TeV pp
- 24 May, 2018 - Update LAr veto tags
- 17 Feb, 2018 - Updated 2017 luminosity available in Tag OflLumi-13TeV-010
Recommendations
Data Sample | Recommended Luminosity Tag | Recommended Livefraction Trigger | Comments |
---|---|---|---|
data18_hi | OflLumi-HI-009 (OflLumi-HI-007 or later) | L1_EM12 | Preliminary 2018 calibration |
data18_900GeV | OflLumi-900GeV-002 | L1_ALFA_ELAST15 | Preliminary 2018 calibration |
data18_13TeV | OflLumi-13TeV-010 | L1_EM24VHI | Preliminary 2018 calibration |
2017/18 13TeV low-mu | OflLumi-13TeV-010-lowmu | L1_EM24VHI | Preliminary 2017/18 low-mu data calibration |
data17_5TeV | OflLumi-5TeV-003 | L1_EM24VHI | Preliminary 2017 5 TeV calibration |
data17_13TeV | OflLumi-13TeV-010 | L1_EM24VHI | Preliminary 2017 calibration |
data16_13TeV | OflLumi-13TeV-010 (OflLumi-13TeV-009 or later) | L1_EM12 | Final 2016 calibration |
data16_hip8TeV | OflLumi-HI-009 (OflLumi-HI-006 or later) | L1_EM12 | Updated calibrations (Winter 2019) |
data16_hip5TeV | OflLumi-HI-009 (OflLumi-HI-004 or later) | L1_EM12 | Online calibration |
data15_13TeV | OflLumi-13TeV-010 (OflLumi-13TeV-004 or later) | L1_EM12 | Final 2015 vdM calibration |
data15_hi | OflLumi-HI-009 | L1_EM12 | Final 2015 vdM estimate |
data15_5TeV | OflLumi-5TeV-003 | L1_EM12 | Final vdM calibration |
data13_hip | OflLumi-HI-003 | L1_LUCID_A_C | Final vdM calibration |
data13_2p76TeV | OflLumi-2p76TeV-002 | L1_LUCID_A_C | Final 2.76 lumi |
data12_8TeV | OflLumi-8TeV-004 | L1_EM30 | Final 2012 vdM calibration |
data11_7TeV | OflLumi-7TeV-004 | L1_EM30 | Final 2011 results (with beam-beam corrections) |
data10_7TeV | OflLumi-7TeV-004 | L1_MBTS_2 | Final 2010 results (with beam-beam corrections) |
data11_2p76 | OflLumi-2p76TeV-002 | L1_LUCID_A_C | Final 2.76 lumi |
data10_hi | OflLumi-HI-000 | L1_LUCID_A_C | Online estimate |
Luminosity Calculator
See below for a synopsis of these quantities.
Definitions
Good Run List
A Good Run List (GRL) is an ATLAS standard XML file format which defines a list of Run and Luminosity Blocks to be considered. The only officially supported source of GRLs is through the Good Runs List Generator. There is also a growing list of pre-computed standard GRLs available.
To specify a subset of an entire GRL, you can use the the run specifier in the Other Options box. For example, -r 180614-180776,182726-183462 will only process runs in the XML file found in that run range. This happens to correspond to periods E and G. You can use a single run, a range, an open-ended range (e.g.: -r 182726-), or combine any of these into a comma separated list. A run which satisfies any of these run-range criteria will be processed.
Luminosity Tag
The luminosity tag specifies the tag of the offline luminosity folder in COOL to use. Details on the available tags can be found at the Offline Luminosity Tags TWiki page.
Live Fraction Trigger
Technically a question of trigger efficiency, for convenience the luminosity calculator also will compute the L1 live fraction, which is the fraction of delivered luminosity which ATLAS actually recorded. Triggered events can be ignored (vetoed) for a number of reasons, including DAQ deadtime, HLT or SFO backpressure, or because the DAQ is paused. Currently, the method for measuring the L1 live fraction is to count the number of triggers before and after veto which is stored independently in COOL. In typical operations, all triggers in a given bunch group should observe an identical veto rate (and hence share the same live fraction). The best trigger to use to make this determination is a trigger which has a high rate after prescale, so that the statistical uncertainty per luminosity block is low.
The recommended live fraction triggers are:
- 2010 pp: L1_MBTS_2 - this was the highest rate L1 trigger in 2010
- 2010 HI: L1_LUCID_A_C - due to problems with the MBTS in HI running, this backup min-bias trigger is preferred
- 2011 pp before September: L1_EM14 - guaranteed to be in the menu with rather high rate, this is the best option for 2011 data until September, removed from the menu in early September
- 2011 pp: L1_EM30 - present for all of 2011, still has adequate rate, this is the best option for the complete 2011 data sample
- 2012 pp: L1_EM30 - present for all of 2012
- 2013: L1_LUCID_A_C - best minbias trigger for both pp at 2.76 TeV and p-Pb.
- 2015: L1_EM12 - highest pT L1 EM trigger
- 2016: L1_EM12 - highest pT L1 EM trigger
- 2017: L1_EM24VHI - highest pT L1 EM trigger
- 2018: L1_EM24VHI - highest pT L1 EM trigger
- 2018, 900 GeV: L1_ALFA_ELAST15 - higher rate trigger for low-mu data
- 2018 HI: L1_EM12
One can also check the livefraction using any L1 physics trigger and similar results should be obtained. Problems can be found if the prescaled trigger has a low enough rate to be in the range of Poisson statistics per lumi block, although summed over many luminosity blocks, the overall livefraction rate should still be accurate.
Physics Trigger
As another convenience, the luminosity calculator can also incorporate any trigger prescale into the definition of luminosity, but similar to the live fraction this is really a measure of the trigger efficiency. It is important to always specify the ultimate EF trigger, as the relevant prescale factor is the product of L1, L2, and EF prescales. Even though a given L1 trigger may be unprescaled, it doesn't mean that the HLT hasn't been configured to throw events away. The luminosity calculation divides the delivered luminosity per lumi block by the total prescale factor to arrive at the luminosity after prescale. The column reporting Lumi-weighted prescale is simply the ratio of total luminosity with and without this correction. If you prefer to deal with the prescales somewhere else in your analysis, simply specify None and a prescale of 1 will be assumed.
If the specified trigger item doesn't exist or is disabled for a given luminosity block (indicated in TrigConf by a prescale of -1), the luminosity calculator will treat this as an infinite prescale. This means this lumi block will not contribute to the delivered luminosity after prescale. In addition, this luminosity block will be marked as 'Bad' in the accounting of Good/Bad luminosity blocks. The other reason why a lumi block will be flagged as bad is if the luminosity value is known to be invalid. No lumi block with a Green LUMI DQ flag should contain invalid data.
Please note that the random trigger rates are not currently accounted for by the luminosity calculator. Any random-seeded trigger will typically have a luminosity reported which is much larger than the true luminosity.
LAr Event Veto
Starting with the Release 17 reprocessing (Summer 2011) and all data taken from September 2011 onwards, certain LAr defects (noise bursts) are not rejected at the GRL stage. Users must reject specific events instead, as more fully described on the LArEventVetoRel17 TWiki. The loss of luminosity from applying these event vetos is calculated by determining what fraction of a given luminosity block has fallen into one of these LAr EventVeto periods. The LAr event veto database is versioned, and one must use the DB tag which matches the version of the data being processed. Please consult the LArEventVetoRel17 TWiki for the latest recommendation. For 2017, the data reconstructed at the Tier0 (t0pro21_v01 containers) needs LARBadChannelsOflEventVeto-RUN2-UPD4-06, while the reprocessed data will use LARBadChannelsOflEventVeto-RUN2-UPD4-08. Either tag is fine for Run2 data before 2017. As of 2018, LARBadChannelsOflEventVeto-RUN2-UPD4-10 contains the best information for all running periods.
If you want to specify a tag not present in the pop-up list, specify none, and in the Other Options box use --lar --lartag=TAG.
Other Options
Additional options can be specified in the text box, or by using the handy check boxes for commonly used options. For triggers which use BGRP7, you must specify channel 17 to get the BGRP7 luminosity. For triggers which include b-jet triggers, you must require the online beamspot status to be valid to get the correct livefraction.